1. Brezizgubno stiskanje rastrskih slik z uporabo genetskega algoritmaTomaž Klobučar, 2025, master's thesis Abstract: V magistrskem delu je predstavljena uporaba genetskega algoritma za brezizgubno stiskanje rastrskih slik. Poudarek je na kombiniranju genetskega algoritma z različnimi tehnikami stiskanja podatkov, vključno z aritmetičnim kodiranjem, metodo RLE (angl. Run Length Encoding) in Huffmanovim kodiranjem. Podrobno je opisano teoretično ozadje genetskega algoritma in njegovih osnovnih postopkov, kot so selekcija, križanje in mutacija. Prav tako je predstavljena implementacija genetskega algoritma, kodirnika in dekodirnika. Opravljene so bile analize vhodnih parametrov kodeka, stiskanja splošnih in risanih slik, vpliva napovedi genetskega algoritma na stopnjo stiskanja, vpliva pretvorbe barvnega prostora na stopnjo stiskanja ter analiza časovne zahtevnosti. Rezultati so pokazali, da predlagan kodek doseže stopnjo stiskanja primerljivo z izbranimi formati, njegova učinkovitost stiskanja pa se izboljša z uporabo pretvorbe barvnega prostora. Keywords: brezizgubno stiskanje slik, risane slike, genetski algoritem, Huffmanovo kodiranje, aritmetično kodiranje, RLE Published in DKUM: 06.02.2025; Views: 0; Downloads: 1
Full text (7,00 MB) |
2. Energy flexibility in aluminium smelting : a long-term feasibility study based on the prospects of electricity load and photovoltaic productionMarko Bizjak, Niko Uremović, Domen Mongus, Primož Sukič, Gorazd Štumberger, Haris Salihagić Hrenko, Dragan Mikša, Stanislav Kores, Niko Lukač, 2024, original scientific article Keywords: energy flexibility, aluminium smelting, renewable energy, virtual battery, solar production Published in DKUM: 17.12.2024; Views: 0; Downloads: 5
Full text (1,91 MB) |
3. Novel GPU-accelerated high-resolution solar potential estimation in urban areas by using a modified diffuse irradiance modelNiko Lukač, Domen Mongus, Borut Žalik, Gorazd Štumberger, Marko Bizjak, 2024, original scientific article Abstract: In the past years various methods have been developed to estimate high-resolution solar potential in urban areas, by simulating solar irradiance over surface models that originate from remote sensing data. In general, this requires discretisation of solar irradiance models that estimate direct, reflective, and diffuse irradiances. The latter is most accurately estimated by an anisotropic model, where the hemispherical sky dome from arbitrary surface’s viewpoint consists of the horizon, the circumsolar and sky regions. Such model can be modified to incorporate the effects of shadowing from obstruction with a view factor for each sky region. However, state-of-the-art using such models for estimating solar potential in urban areas, only considers the sky view factor, and not circumsolar view factor, due to high computational load. In this paper, a novel parallelisation of solar potential estimation is proposed by using General Purpose computing on Graphics Processing Units (GPGPU). Modified anisotropic Perez model is used by considering diffuse shadowing with all three sky view factors. Moreover, we provide validation based on sensitivity analysis of the method’s accuracy with independent meteorological measurements, by changing circumsolar sky region’s half-angle and resolution of the hemispherical sky dome. Finally, the presented method using GPPGU was compared to multithreaded Central Processing Unit (CPU) approach, where on average a 70x computational speedup was achieved. Finally, the proposed method was applied over a urban area, obtained from Light Detection And Ranging (LiDAR) data, where the computation of solar potential was performed in a reasonable time. Keywords: solar energy, solar potential, anisotropic diffuse irradiance, LiDAR, GPGPU Published in DKUM: 17.12.2024; Views: 0; Downloads: 3
Full text (8,06 MB) |
4. |
5. Investigating the influence of reflective materials on indoor thermal environment and solar reflectance in buildingsJihui Yuan, Yasuhiro Shimazaki, Masaki Tajima, Shaoyu Sheng, Zhichao Jiao, Marko Bizjak, 2024, original scientific article Abstract: This in-depth study explores the intricate dynamics of reflective materials, emphasizing their impact on the indoor thermal environment and urban heat island (UHI) mitigation. Examining diffuse highly reflective (DHR), general reflective (GR), and retro-reflective (RR) materials on a simplified building model during summer days, the research utilizes outdoor experiments to analyze air and surface temperatures, as well as solar radiation. Prioritizing key metrics—mean radiant temperature (MRT), operative temperature (OT), and solar reflectance (ρ)—the study uncovers nuanced distinctions in DHR, GR, and RR materials. Solar reflectance calculations consistently show higher values for DHR and RR materials compared to GR material, highlighting reflectance's pivotal role in influencing surface temperatures and indoor thermal environment. When evaluating the impact of exterior wall materials on building temperatures, RR material with a 76% reflectance performs similarly to DHR material (82%). Notably, with a 6% lower reflectance in RR, the temperature contrast between external and internal walls is only about 1.5 °C at its maximum, underscoring RR's effectiveness as an outer wall material for UHI mitigation and building energy conservation, surpassing both DHR and GR materials. Keywords: urban heat island, reflective materials, indoor thermal environment, solar reflectance, simplified building model Published in DKUM: 23.08.2024; Views: 110; Downloads: 9
Full text (4,82 MB) |
6. Poučevanje računalniških omrežij s pomočjo obogatene resničnosti : magistrsko deloAleš Spital, 2024, master's thesis Abstract: Računalniška omrežja imajo ključno vlogo v učnem načrtu srednjega tehniškega izobraževalnega programa tehnik računalništva. Današnje strategije poučevanja računalniških omrežij se pogosto osredotočajo na teoretična predavanja, kar lahko predstavlja izziv za učence pri razumevanju, kako se to znanje uporablja v praksi. Zato smo v zaključnem delu preučili možnosti vključevanja obogatene resničnosti in razvili interaktivno mobilno aplikacijo. Ta omogoča dijakom virtualno raziskovanje in konfiguriranje omrežnih komponent ter spremljanje prometa pošiljanja paketov. Na koncu smo izvedli študijo z dijaki, ki je ocenila vpliv izdelane mobilne aplikacije na razumevanje računalniških omrežij v primerjavi s tradicionalnimi metodami učenja. Keywords: poučevanje, obogatena resničnost, umetna inteligenca, mobilna aplikacija, računalniška omrežja Published in DKUM: 01.07.2024; Views: 123; Downloads: 26
Full text (4,09 MB) |
7. Novel Half-Spaces Based 3D Building Reconstruction Using Airborne LiDAR DataMarko Bizjak, Domen Mongus, Borut Žalik, Niko Lukač, 2023, original scientific article Abstract: Automatic building reconstruction from laser-scanned data remains a challenging research topic due to buildings’ roof complexity and sparse data. A novel automatic building reconstruction methodology, based on half-spaces and a height jump analysis, is presented in this paper. The proposed methodology is performed in three stages. During the preprocessing stage, the classified input point cloud is clustered by position to obtain building point sets, which are then evaluated to obtain half-spaces and detect height jumps. Half-spaces represent the fundamental shape for generating building models, and their definition is obtained from the corresponding segment of points that describe an individual planar surface. The detection of height jumps is based on a DBSCAN search within a custom search space. During the second stage, the building point sets are divided into sub-buildings in such a way that their roofs do not contain height jumps. The concept of sub-buildings without height jumps is introduced to break down the complex building models with height jumps into smaller parts, where shaping with half-spaces can be applied accurately. Finally, the sub-buildings are reconstructed separately with the corresponding half-spaces and then joined back together to form a complete building model. In the experiments, the methodology’s performance was demonstrated on a large scale and validated on an ISPRS benchmark dataset, where an RMSE of 0.29 m was obtained in terms of the height difference. Keywords: LiDAR point cloud, building reconstruction, half-spaces, Boolean operations Published in DKUM: 01.12.2023; Views: 388; Downloads: 25
Full text (13,79 MB) This document has many files! More... |
8. Razvoj napovednega modela multivariatnih časovnih vrst uporabniških storitev : diplomsko deloSandi Pečečnik, 2023, undergraduate thesis Abstract: V sklopu diplomskega dela predstavimo več nevronskih mrež, ki jih optimiziramo, pri čemer raziščemo ustrezne arhitekture, metrike, funkcije in druge pomembne lastnosti za uporabo v napovednih modelih multivariantnih časovnih vrst. Raziščemo najpomembnejše probleme povezane z razvojem napovednih nevronskih mrež. Naslovimo reševanje treh specifičnih realnih problemov, za reševanje katerih smo predlagali arhitekture nevronskih mrež. Izdelali smo še skalabilno spletno aplikacijo, ki omogoča enostavnejšo uporabo naučenih modelov nevronskih mrež. Keywords: časovne vrste, nevronske mreže, globoko učenje, storitve, arhitekture globokega učenja Published in DKUM: 05.10.2023; Views: 453; Downloads: 46
Full text (1,66 MB) |
9. |
10. Spletni koledar s samodejnim predlaganjem skupinskega prostega termina : diplomsko deloTilen Heric, 2022, undergraduate thesis Abstract: Zaradi problema organiziranja svojega časa poleg velikega števila dogodkov smo z namenom olajšanja organiziranja le teh ustvarili spletni koledar z možnostjo samodejnega predlaganja termina. Namen spletne aplikacije je olajšati organizacijo dogodkov skupinam, saj pri velikem številu članov predstavlja organiziranje dogodkov, ki bi se ga lahko udeležili vsi člani, velik problem. Cilj spletne aplikacije je s pomočjo algoritma podati uporabniku čim boljši termin zanj ali za njegovo skupino. Implementirali smo zaledni in prednji del spletne aplikacije ter vključili uporabo dveh algoritmov. Pri zalednem delu smo uporabili MongoDB za shranjevanje podatkov, Node.js kot strežnik in ogrodje Express po vzorcu MVC. Za prednji del aplikacije smo uporabili ogrodje React. Zaledni in prednji del komunicirata preko klicev Rest API. Uporabniku smo ob ustvarjanju novega dogodka z uporabo algoritma grobe sile izpisali proste termine in z uporabo algoritma Simuliranega ohlajanja predlagali termin. V rezultatih smo ugotovili, da algoritem grobe sile deluje dovolj hitro za interaktivno upravljanje z aplikacijo. Algoritem Simulirano ohlajanje omogoča vnašanje dodatnih kriterijev in deluje še vedno dovolj hitro za interaktivno upravljanje z aplikacijo. Keywords: spletni koledar, Simulirano ohlajanje, predlaganje prostega termina Published in DKUM: 24.10.2022; Views: 500; Downloads: 45
Full text (1,75 MB) |